The primary focus of this subproject will be the molecular biology of adeno-associated virus as it relates to the use of AAV for gene therapy. AAV has shown significant promise as a vector for therapy in a variety of tissues and one of the most promising targets is the nervous system. Nevertheless, significant basic and technical issues need to be resolved before rAAV vectors can be used with confidence. One of these is the role of the virus encoded Rep protein in viral DNA replication, integration and control of gene expression. To address this issue, the first goal of this subproject will be to isolate conditional lethal mutations in the AAV rep gene. These mutants will be used to define more precisely the domains of the Rep protein involved in the DNA replication, activation and repression functions of Rep. In addition, these mutants will be used to answer questions about the role of Rep in site specific integration into human chromosome 19. The second goal of this subproject will be to improve the current methods of rAAV virus production. Previous attempts to develop packaging lines or adenovirus/AAV hybrid viruses have apparently floundered because of the toxicity observed when the AAV rep genes are expressed in human cells at high copy number. We will focus on the development of adenovirus vectors which carry AAV genes in a non-rescuable form. Several methods will be tried to engineer stable adenovirus/AAV hybrid vectors which can be used to provide efficiently all of the Ad and AAV gene functions required for rAAV propagation. These will include the use of the conditional lethal rep mutations proposed above, the use of inducible rep promoters, and the use of novel cell lines that down regulate rep expression. The last goal of this subproject will be to determine the efficiency and persistence of rAAV transduction in primary neural tissues, in vivo, specifically the spinal cord, as a function of time and the promoter used to drive expression of the transgene. rAAV virus carrying an improved green fluorescent protein (gfp) gene will be injected into normal rat nervous tissue in vivo. The gfp marker gene will be under the control of neural specific, glial specific or housekeeping promoters. Sites of injection will include spinal cord, hippocampus, cortex an cerebellum. Expression at the sites of injection will be monitored by several methods as a function of time over the course of ear. The physical state of the viral genome, that is, whether it exists as an episome or as an integrated provirus, will also be determined. These experiments will complement similar studies in retinal tissues proposed in Subproject 2, and provide information for the studies proposed in Subproject 3, in which rAAV will be used for the expression of specific therapeutic transgenes in the spinal cord.